Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1106834 | Transportation Research Procedia | 2015 | 11 Pages |
The availability of floating car data (FCD) enables operators to use novel methods in travel time estimation. A first step towards combining traffic data from loops and FCD is to check the compatibility between the two types of travel time estimates. We perform an in-depth statistical analysis that allows us to compare various travel time estimates using data collected from the peri-urban highways in the region of Lille, in north France. The comparison is performed separately for light and heavy vehicles and for various settings: peak hour, off-peak hours, working day, holiday, rain, and so on. The results show that the two estimates are linearly correlated and a specific function can be calibrated for each site for itineraries of variable length. Overall, this paper provides evidence that different flow regimes necessitate differentiated a priori treatment in order to enhance the reliability of estimates made on data coming from different sources.